IBM Calls for Patent Reform 292
daria42 writes "IBM has called for tighter regulation of patents and a review of intellectual property ownership issues in collaborative software development. The company is one of the largest patent-holders in the United States. IBM executive Jim Stallings said examining patents for prior art should not only be the job of the patent office but that the wider community should be involved. Stallings also called on the industry to stop what he calls "bad behaviour" by companies who either seek patents for unoriginal work or collect and hoard patents."
Interesting insights by "innocent IBM"... ;-/ (Score:5, Informative)
It's good to see how the taste of their own medicine in cases such as the SCO litigation finally seems to lead IBM back to their initial stance of speaking out against software patents - from one of the world's largest patent holders, the obligation to "use or lose IP" as in trademark law is quite a remarkable one.
IBM is Similar to Microsoft (Score:3, Informative)
Re:Actual research by the patent office? (Score:3, Informative)
Not just software! (Score:4, Informative)
Apparantly people who make Real Things(tm) have problems with patents too!
Response will likely be useless (Score:3, Informative)
Since the late 1960s the management culture at the PTO has shifted from a strong "nothing is patentable unless you really, really make a good argument and present narrow claims" to a "keep application pendency down by getting allowances/abandonments asap" approach. Since the applicant can argue rejections, which take time and energy to rebut in a new rejection, it is easier to just say "I agree" and allow the application. This is enforced by a quota system which, simplified, amounts to counting disposals (allowance by the examiner, abandonment by the applicant, or writing an examiner's answer to an appeal to the Board of Appeals filed by the applicant). When you throw in other performance elements such as responding within set time limits to applicant's responses, starting work on the oldest application in an examiner's docket, the actual "quality" performance elements (finding the best prior art and applying in logical, well reasoned rejections) are actually a relative minority of an examiner's performance rating (looking at page 35 of the following Inspector General's Report [doc.gov], the first three elements are "quality elements, amounting to 40%; the remaining 60% are production time elements, service to the public, etc. This report, incidentally, is even more PHB oriented than the incredible level that PTO management already is)
Added to this mix is that fact that, although the time allotted to examine each application is roughly unchanged over the last few decades it has gotten tougher to do a quality examination than it was years ago. This has been caused by several factors, but the main ones are:
less ability to "write off" time spent on important examining related activities, such as maintaining search files (classifying foreign patents and the mass of non-patent documents into the Patent Classification scheme, which amounts to a way of "tagging" such documents, a valuable supplement to full text searching, and historically, the primary means to search)
ever increasing requirements to justify making rejections; they have to discuss in bloody detail how each element of the claim is shown by the prior art, used in the same way, and, for obvious rejections, why it is proper to used disclosures from two or more references, the "motivation" to do so having to be derived by statements actually in the references, and not just by the examiner's deductive reasoning. Indeed, a rejection can be judged to be improper for not being properly supported by the cited prior art, counting as an error against the examiner.
lots of other, constant distractions, procedures, and requirements encountered on a daily basis that have only multiplied over the years.
When management is confronted with the complaints from influential "patent system users" such as IBM here, their typical response is to institute "quality review" programs where people who can't be as expert as the typical examiner who regularly works in the art of the application being reviewed is making judgements on issues like search quality, relevance of prior art to the claims, etc. The reviewers can do additional searches, basically spending more search time on the application, but with the consequence that the examiner will be charged with an error and suffer a decreased performance rating, rather than having more time to search the application and find the relevant art in the first case. What PTO management tries to do is like what PHBs in programming shops try to do by demanding more bug free code by increasing the QA department and then firing programmers who wrote the buggy code without changing the ever shorter deadlines. And, remember, that patent examination is not project work, but production work, that just on and an with no breaks until retirement, resignation, firing, or death.
I work for IBM... (Score:5, Informative)
It is harder to get a patent through the IBM internal review process (which you have to complete before Legal will consider filing it) than it is to get the USPTO to take it. A team of other IBMers (usually inventors themselves) ensure that it is innovative and possible. The revenue possibilities are also considered.
I've put 11 ideas through the process, all of which I am reasonably sure would have been granted by the USPTO, and only 1 of those has been applied for. Three were thought worth protecting but not patenting -- they are published on the ip.com prior art database to prevent anyone else patenting them.
If all corporations handled patents as responsibly as IBM we wouldn't have a problem.